基于非线性重构模型的植物叶片图像集分类方法

Plant Leaf Image Set Classification Approach Based on Non-linear Reconstruction Models

提出一种基于非线性重构模型的植物叶片图像集的分类识别方法。该方法首先使用高斯受限玻尔兹曼机(GRBMs)通过非监督预训练来初始化模型的权值;然后针对每一个植物叶片图像集用初始化的模型训练得到一个特定的模型;最后根据测试样本的最小重构误差和测试样本集的最多投票策略来判定测试样本集的类别。该方法通过图像预处理来处理图像,避免了图像在缩放时发生形变,并采用基于k-means的特征提取方法来提取植物叶片图像特征。实验结果表明,该方法能够准确地对植物叶片图像集进行分类识别。

In this paper, a plant leaf image set identification approach was proposed based on non-linear reconstruction models. This approach initializes the parameters of model by performing unsupervised pre-training using Gaussian re-stricted Boltzmann machines（GRBMs）. Then, the pre-initialized model is separately trained for images of each plant set and class-specific models are learnt. At last, based on the minimum reconstruction error from the learnt class-specific models,majority voting strategy is used for classification. Besides, in order to avoid occurring deformation during the image scaled, this paper normalized plant image by image preprocessing and a method of feature extraction was used based on k-means. The experimental results show that this approach can accurately classify the class of plant image set.